EDR: Postgres service failed to start due to PANIC replication checkpoint has wrong magic 0

Products

• Carbon Black EDR (formerly Cb Response)

Issue/Introduction

PANIC reported in /var/log/cb/pgsql/postgres.log

2021-11-09 14:01:54 +04 [(6463) @ ] LOG:  startup process (PID 6465) was terminated by signal 6: Aborted
2021-11-09 14:01:54 +04 [(6463) @ ] LOG:  aborting startup due to startup process failure
2021-11-09 14:01:54 +04 [(6463) @ ] LOG:  database system is shut down
2021-11-09 14:01:57 +04 [(6468) @ ] LOG:  database system was interrupted; last known up at 2021-10-14 01:24:55 +04
2021-11-09 14:01:58 +04 [(6468) @ ] PANIC:  replication checkpoint has wrong magic 0 instead of 307747550

 

Environment

• EDR(Formerly Carbon Black Response) Server: All Supported Versions

Cause

Corruption in "pgsql/data/pg_logical/replorigin_checkpoint" file

Resolution

• Move out the "pgsql/data/pg_logical/replorigin_checkpoint" file to somewhere else for backup purpose.
• Restart the PostgreSQL database or Restart EDR services

7.4 and Above: 
/usr/share/cb/cbservice cb-pgsql restart

7.3 and Below: 
service cb-pgsql restart